Title: Assessing the performance, usability and cognitive workload of an AI-based navigation assistant: a quantitative study with people with visual impairments
Authors: Bineeth Kuriakose; Raju Shrestha; Frode Eika Sandnes
Addresses: Department of Computer Science, Oslo Metropolitan University, Oslo, Norway ' Department of Computer Science, Oslo Metropolitan University, Oslo, Norway ' Department of Computer Science, Oslo Metropolitan University, Oslo, Norway
Abstract: Recent advances in artificial intelligence (AI) have propelled the development of assistive navigation systems for people with visual impairments. Many studies focusing on the benefits of using AI in navigation assistants overlook user perceptions and responses. Human factors are crucial in these systems, aiming to demonstrate enhanced capabilities and gain user acceptance. Notably, involving users in evaluations is vital for a realistic understanding of system usability and performance - a critical aspect frequently neglected in existing systems. This paper presents results from a study assessing an AI-based smartphone navigation assistant for the visually impaired. In this quantitative study, 13 users evaluated the navigation assistant in natural environments. Results indicate that although the navigation assistant helps in avoiding obstacles, participants express reservations about relying solely on an AI navigation assistant. These findings emphasise the need for a user-centric approach in designing and implementing AI technologies to improve user acceptance and usability.
Keywords: performance; usability; cognitive workload; quantitative study; artificial intelligence; AI; smartphone; navigation assistant; blind; visual impairments; assistive technology; user perceptions; human factors; user evaluation; user acceptance; deep learning; portable; DeepNAVI.
DOI: 10.1504/IJHFE.2024.139166
International Journal of Human Factors and Ergonomics, 2024 Vol.11 No.2, pp.129 - 156
Received: 10 May 2023
Accepted: 20 Sep 2023
Published online: 24 Jun 2024 *